Boosting Multi-Label Image Classification with Complementary Parallel Self-Distillation

Authors: Jiazhi Xu, Sheng Huang, Fengtao Zhou, Luwen Huangfu, Daniel Zeng, Bo Liu

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive results on MS-COCO and NUS-WIDE datasets demonstrate that our framework can be easily plugged into many MLIC approaches and improve performances of recent stateof-the-art approaches.
Researcher Affiliation Collaboration 1School of Big Data and Software Engineering, Chongqing University 2Fowler College of Business & CHDMA, San Diego State University, 3Institute of Automation, Chinese Academy of Sciences, 4JD Finance America Corporation
Pseudocode No The paper describes the methodology using text and mathematical equations but does not include any structured pseudocode or algorithm blocks.
Open Source Code Yes The source code is released at https://github.com/Robbie-Xu/CPSD.
Open Datasets Yes Two widely used MLIC datasets, named MS-COCO and NUS-WIDE, are used for the evaluation of our method. MS-COCO contains 122,218 images with 80 categories of objects in natural scenes, including 82,081 images for training and 40,137 images for validation. In the official partition of NUS-WIDE dataset, it contains 125,449 labeled training pictures and 83,898 labeled test pictures from Flickr, which share 81 labels in total.
Dataset Splits Yes MS-COCO contains 122,218 images with 80 categories of objects in natural scenes, including 82,081 images for training and 40,137 images for validation.
Hardware Specification No The paper does not explicitly describe the specific hardware (e.g., GPU models, CPU types, or cloud instances) used for running the experiments.
Software Dependencies No The paper mentions software components like 'Adam optimizer' and 'ASL' and model types like 'ResNet101-TF' and 'GloVe', but it does not specify version numbers for any software dependencies.
Experiment Setup Yes All experiments follow a training pipeline where Adam optimizer is used with weight decay of 10-4 under a batch size of 32. ASL is applied as the default classification loss function, and the hyper-parameters of ASL are simply left as their default settings. τ in Equation 1 is set to be 3. We set the training epoch to be 20 and 80 for sub-models and the compact global model individually.